diff --git a/nemo/lightning/pytorch/strategies.py b/nemo/lightning/pytorch/strategies.py
index 57cd33a612ae..a17bdd60c77c 100644
--- a/nemo/lightning/pytorch/strategies.py
+++ b/nemo/lightning/pytorch/strategies.py
@@ -441,7 +441,9 @@ def validation_step(self, dataloader_iter, *args: Any, **kwargs: Any) -> STEP_OU
         kwargs = self._update_step_kwargs(dataloader_iter, kwargs, "validation")
 
         with self.precision_plugin.val_step_context():  # TODO: Do we need this?
-            return self.model(dataloader_iter, forward_only=True, *args, **kwargs)
+            out = self.model(dataloader_iter, forward_only=True, *args, **kwargs)
+            self.lightning_module.log('val_loss', out, rank_zero_only=True, batch_size=1)
+            return out
 
     @override
     def test_step(self, dataloader_iter, *args: Any, **kwargs: Any) -> STEP_OUTPUT: